--- Day changed --- Log opened Sat May 31 00:00:01 2003 01:10 -!- lemon [~lemon@beijing.ddns.comp.nus.edu.sg] has joined #uml 02:55 < lemon> green: hi, remeber we discussed 'redirection of uml boot mesg' last time? : ) i've got one solution. 02:57 < lemon> green: i tried manaully './linux > boot_mesg'. it worked. everything from boot goes to the file. also data thrown into /dev/console and /dev/vc/0 also goes there. ; ) 03:03 -!- mace [mace@bub.ipv6.darksilence.net] has quit [Ping timeout: 492 seconds] 03:06 -!- mace [~mace@bub.ipv6.darksilence.net] has joined #uml 03:59 < green> lemon: cool 04:00 < lemon> green: hi, i've been working on it for a while since my last mesg. there are new observations i'd like to share with you. 04:02 < lemon> the redirection of stdout works fine as long as i don't do any redirection of stdin, like what i did yesterday, i.e. 'open(STDIN, "< /dev/null")' 04:03 < green> ah 04:03 < green> hm. interesting 04:03 < lemon> strange. ; ) 04:04 < lemon> also noticed from 'less -f boot_mesg' is that there are additional '^M' shown. 04:05 < green> this is added by the uml itself most probably 04:05 < lemon> yesterday, when the redirection problem was not solved, there were no such '^M' from the partial boot_mesg. 04:05 < lemon> possible. 04:06 < green> well, at least when I get kernel output from real kernel via serial console, I always get these extra ^M ;) 04:06 < lemon> i see. i've never tried a serial console before. cool 04:09 < lemon> also now stdin of the parent process, which forked the uml binary, seemed block. (just like when i start uml manually from the shell, the shell blocks every keyboard input including ctrl-c) 04:09 < green> sure, because you have not redirected it 04:11 < lemon> you mean the uml binary take control of the stdin? 04:11 < green> yes 04:13 < lemon> please forgive my ignorance. so a forked process can not only inherits its parents file descriptor but also make modifications (like block) to it, whcih can be seen by the parent process? 04:14 < lemon> i am not very sure about I/O in shell. ; ) 04:16 < green> well, it should not block it, but it van intercept it for sure 04:16 < green> e.g do cat & ;) 04:20 < lemon> i think i am getting it. so what i typed goes to uml, the forked process, not its parent. 04:21 < lemon> could there be contention on stdin between the parent and its child process? 04:26 < green> yes 04:27 < lemon> i see. : ) 04:38 < lemon> green: hi, one more question i'd like to ask. after the the parent process forked uml, whatever message printed from the parent process (the uml controller) does not seem inlined ( it seems addition '\t' are added by the system). 04:39 < lemon> i print out the $ENV{'TERM'} from the parent process. it says 'screen'. I am using 'screen' right now. 04:39 < lemon> any idea? 04:44 < green> no, I am not even understand what you are asking about ;) 04:50 < lemon> say 'print("\thello". "\n" . "\thello2\n"). the output should be two line, both of with have a tab in front of them. 04:51 < lemon> the problem here is that in the output of the 2nd line, there is an additional tab charactro added. 04:52 < lemon> say there is a print("\thello3\n") following the pre. statement. 04:53 < lemon> the output is that there are 3 tabs all together before 'hello3'. : ) 04:55 < lemon> got it? : ) any ideas why there are those additional tab chars? 04:56 < lemon> opps. library is closing down. thanks for your help. see you. : ) 04:56 -!- lemon [~lemon@beijing.ddns.comp.nus.edu.sg] has quit [Quit: Leaving] 07:17 -!- Getty [~Getty@metaluna4.de] has quit [Ping timeout: 492 seconds] 07:29 -!- Getty [~Getty@metaluna4.de] has joined #uml 09:03 -!- tp [~posch@130.149.49.70] has joined #uml 09:18 -!- tp [~posch@130.149.49.70] has left #uml [Client Exiting] 10:03 -!- Lathiat_ [lathiat@130.95.13.25] has joined #uml 10:03 -!- Lathiat_ [lathiat@130.95.13.25] has quit [Client Quit] 10:11 < Lathiat> Anyone alive? 10:11 < Lathiat> wondering if someone knows how to start a screen session withotu actually attaching to it 10:18 < mistik1> screen -a -m 10:18 < mistik1> oops 10:18 < mistik1> screen -d -m 10:18 < mistik1> starts it in detached mode 10:21 < Lathiat> thanks 10:22 < mistik1> np 10:42 < Lathiat> hi guys anyoen alive 10:42 < Lathiat> + .long SYMBOL_NAME(sys_new_s_context) 10:42 < Lathiat> + .long SYMBOL_NAME(sys_set_ipv4root) 10:42 < Lathiat> thts from arch/i386/entry.S 10:42 < Lathiat> anyone know what the equivalent to put in arch/um is 10:42 < Lathiat> im guessing its arch/um/kernel/sys_call_table.c 10:43 * mistik1 is no kernel hacker 10:44 < Lathiat> damn 10:44 < Lathiat> ill just guess 10:44 < Lathiat> and if ti dont work 10:44 < Lathiat> oh well 10:44 < Lathiat> only 5mins of compile time lost 10:44 < mistik1> heh 10:46 -!- mace [~mace@bub.ipv6.darksilence.net] has quit [Ping timeout: 492 seconds] 10:55 -!- tp [~posch@130.149.49.70] has joined #uml 10:56 < tp> hi all 10:56 < tp> got a question: #1 0xa00af708 in segv (address=0, ip=0, is_write=0, is_user=0, sc=0xa2660278) 10:57 < tp> shouldn't there usually be a non-null address in ip ? 10:59 < Lathiat> whats that crashing on 10:59 < tp> 2.4.20-3um 11:00 < Lathiat> i mean like 11:00 < Lathiat> what code section 11:01 < tp> you mean cs ? 11:01 < tp> eip = 0, cs = 35 11:03 < tp> i find this quite strange, most register values are 0 11:04 < tp> and i have no clue how to find the crashing instruction without a useful call trace 11:06 < tp> also, i would have expected cs to be 0x10, as usual 11:09 < Lathiat> not sure sorry 11:10 < tp> can i paste the full stack trace somewhere, so you can take a look ? 11:15 < Lathiat> try the mailing list 11:15 < Lathiat> sorry i probably wouldnt be much help 11:15 < tp> it's a pity, i really have no clue where to start searching in this case 11:23 -!- pierre [~pierre@67.34.240.49] has joined #uml 11:24 -!- pierre [~pierre@67.34.240.49] has quit [Client Quit] 12:15 < Lathiat> anyone alive yet? 12:15 < Lathiat> heh 12:17 < tp> yure 12:17 < tp> sure, i mean 12:19 < Lathiat> know much about UMl code? 12:20 < Lathiat> im trying to get a virtualserver patch working with uml 12:20 < Lathiat> need to setup the syscalls 12:21 < tp> i never tried creating syscalls 12:22 < tp> but heard, that it's easy 12:22 < tp> there's a howto somewhere 12:22 < Lathiat> well 12:22 < Lathiat> its more UML-specifically 12:22 < Lathiat> under arc/i386 12:22 < Lathiat> u have like 12:22 < Lathiat> + .long SYMBOL_NAME(sys_new_s_context) 12:22 < Lathiat> + .long SYMBOL_NAME(sys_set_ipv4root) 12:22 < Lathiat> not sure how to do that in arch/um 12:26 < tp> did you have a look at arch/um/kernel/sys_call_table.c 12:27 < tp> i guess, your patch modified arch/i386/entry.S ? 12:28 < Lathiat> yeh 12:29 < Lathiat> it does 12:29 < Lathiat> i loooked at the file 12:29 < Lathiat> maybe i added a wrong bit 12:29 < Lathiat> where would i be looking 12:29 < Lathiat> to add those syscalls 12:29 < tp> you got the numbers for the new syscalls, i.e. __NR_* constants ? 12:30 < Lathiat> ahh 12:30 < Lathiat> i think its 257 and 258 12:30 < tp> have you added them to arch/um/kernel/sys_call_table.c ? 12:30 < Lathiat> i think i added the wrong thing ;) 12:30 < Lathiat> [ __NR_sched_setaffinity ] = sys_ni_syscall, 12:30 < Lathiat> [ __NR_sched_getaffinity ] = sys_ni_syscall, 12:30 < Lathiat> add them after that? 12:31 < tp> looks good to me 12:32 < Lathiat> hrm 12:32 < Lathiat> i wonder where the __NR are set 12:33 < Lathiat> hrm 12:33 < Lathiat> where would i define the __NR thing 12:33 < tp> posch@asterix:/mnt/hd/uml/shared/linux/include$ grep -r __NR_ * | less 12:33 < tp> there you find the definitions 12:33 < tp> but your patch probably modifies one of the header files to add the NR constants 12:34 < Lathiat> btw its 253 and 254 12:34 < Lathiat> i think 12:34 < Lathiat> .long SYMBOL_NAME(sys_ni_syscall) /* 250 sys_alloc_hugepages */ 12:34 < Lathiat> .long SYMBOL_NAME(sys_ni_syscall) /* sys_free_hugepages */ 12:34 < Lathiat> .long SYMBOL_NAME(sys_ni_syscall) /* sys_exit_group */ 12:34 < Lathiat> .long SYMBOL_NAME(sys_new_s_context) 12:34 < Lathiat> .long SYMBOL_NAME(sys_set_ipv4root) 12:34 < Lathiat> so 250... 251..252..253 12:34 < tp> where is the patch from ? 12:34 < Lathiat> http://www.solucorp.qc.ca/miscprj/s_context.hc 12:34 < Lathiat> apt-get install vserver if your running debian sid 12:34 < tp> no debian here 12:35 < Lathiat> your havent been turned to the dark side yet *G* 12:36 < tp> he 12:43 < Lathiat> thanks for your help 12:43 < Lathiat> i think i can figure it out now 12:43 < tp> does it work ? 12:43 < tp> ok 12:43 < tp> de rien 12:43 < Lathiat> dunno yet 13:09 < caker> any cisco heads know how to bring down/up an interface? 13:10 * BB used trial and error last time :) 13:10 < caker> well, there's only 1000 commands :-) 13:10 < BB> heh 13:11 < BB> set interface disable ? 13:21 < BB> restart reboot - works on this AT ar320 but probably isnt what you want 13:21 < caker> I found a way to mark it down administrativly via the GUI, but it didn't help the box connected - it has to be restarted :( 13:22 < BB> whats a gui ? lol 13:22 < caker> I don't think this gear has a very solid TCP implementation, since I have to reboot it every few days (it's for remote console access) 13:22 * caker is a cisco newb 13:23 < BB> heh i just hack around until it works, luckily for the lack of knowledge i have we no longer do tech support for cisco stuff ;) 13:23 < caker> hehe who is "we" 13:24 < BB> my place of work ;) heh 13:24 < caker> nice 13:24 < BB> http://www.c-pac.co.uk/ but the work is more instersting than the old website makes it 13:25 < caker> BB: cool! 13:29 < BB> argh my remote is buggered everytime i change channel it turns cable box off :( 13:32 < caker> BB: it's telling you to stop watching tv 13:32 < BB> caker it cant be true, the amount of money i pay for it lol 13:40 < BB> freaky was flat batts 13:52 -!- ElectricElf [david@elf.noc.oftc.net] has quit [Quit: Reboot] 14:07 -!- tp [~posch@130.149.49.70] has quit [Quit: Client Exiting] 14:59 -!- ElectricElf [david@elf.noc.oftc.net] has joined #uml 16:44 -!- karm [karm@62.212.96.119] has joined #uml 17:12 -!- radical [~radical@202.88.184.39] has joined #uml 17:12 -!- radical [~radical@202.88.184.39] has quit [Read error: Connection reset by peer] 17:16 -!- radical [~radical@202.88.184.39] has joined #uml 18:09 -!- radical [~radical@202.88.184.39] has quit [Quit: Client Exiting] 18:46 < david> hello 18:46 < caker> hello david 19:09 -!- slh [~shedii@68.154.227.161] has quit [Remote host closed the connection] 19:11 < caker> I've been hacking my UML kernels to "poweroff" instead of rebooting. 19:11 < caker> and wondering if it's difficult to add an UML kernel switch to do this optionally 19:12 < caker> david: I wanted to make your kernels available on linode.com, but I'd need them to be patched.. 19:12 < caker> i suppose I could use your diffs 19:35 < david> caker: to add a switch you just do a __uml_something 19:35 < david> I can't remember 19:36 < david> caker: I have diffs, or you can send your patch and I can make a new diff 19:36 < caker> I think my patch would be useful if it didn't hard-code it to poweroff on reboot .. a switch I think would be nice 19:37 * caker is no kernel hacker 19:37 < caker> yet 19:37 < david> caker: or just add another mconsole command 19:37 < caker> BTW, I'm starting a two-week beta test on Linode.com tomorrow, if anyone is interested hit me up here or email me (link is on the site) 19:38 < david> caker: sure :-) 19:38 < caker> and the mconsole command would do what? 19:38 < caker> set a bit? 19:38 < david> caker: power it down 19:38 < caker> well, i didn't like the idea of leaving it up to inittab 19:38 < david> nod 19:39 < caker> david: regarding beta testing, just use four and 15 ones as the visa number, and you're set. 19:39 < david> caker: ok, cool 19:39 < david> 41111111111111111111? 19:39 * caker crosses fingers 19:39 < caker> yes 19:39 < david> (didn't count them) 19:39 < david> caker: cool, thanks 19:39 < david> caker: mail me your patch and I can take a peek at it 19:39 < david> got to go work on house now 19:39 < david> bbiab 19:39 < caker> later 19:51 -!- ElectricElf [david@elf.noc.oftc.net] has quit [jupiter.oftc.net ozone.oftc.net] 19:51 -!- ElectricElf [david@elf.noc.oftc.net] has joined #uml 19:56 -!- ElectricElf [david@elf.noc.oftc.net] has quit [ozone.oftc.net ionosphere.oftc.net] 19:56 -!- ElectricElf [david@elf.noc.oftc.net] has joined #uml 20:03 -!- ElectricElf [david@elf.noc.oftc.net] has quit [jupiter.oftc.net ozone.oftc.net] 20:04 -!- ElectricElf [david@elf.noc.oftc.net] has joined #uml 22:51 -!- ASY- [~someone@216.221.51.54] has joined #uml 22:51 -!- ASY [~someone@216.221.51.54] has quit [Read error: Connection reset by peer] 22:51 -!- ASY- is now known as ASY 23:10 -!- pahan [pahan@syr-24-59-98-110.twcny.rr.com] has joined #uml --- Log closed Sun Jun 01 00:00:00 2003